HomeLarge Type Edition
HOME ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
PubMed
Right arrow PubMed Citation
The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 62:997-1001 (2007)
© 2007 The Gerontological Society of America

Limb Fat to Trunk Fat Ratio in Elderly Persons Is a Strong Determinant of Insulin Resistance and Adiponectin Levels

Shai Gavi, Joshua J. Feiner, Mark M. Melendez, Dennis C. Mynarcik, Marie C. Gelato and Margaret A. McNurlan

Departments of 1 Medicine and 2 Surgery, State University of New York at Stony Brook.

Address correspondence to Shai Gavi, MPH, Department of Medicine, Division of General Medicine and Geriatrics, School of Medicine, Health Sciences Center, State University of New York at Stony Brook, Stony Brook, NY 11794-8154. E-mail: sgavi{at}notes.cc.sunysb.edu


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Background. Similar to lipodystrophy syndromes, aging results in increased visceral adiposity with loss of subcutaneous adipose tissue in the extremities. The hypothesis of this study is that the distribution of limb fat to trunk fat (LF/TF) ratio in elderly persons has a stronger correlation than trunk fat alone to insulin resistance and adiponectin levels.

Methods. Thirty-eight elderly participants were divided into an insulin-resistant (IR) group and an insulin-sensitive (IS) group. Limb fat and trunk fat were measured by dual-energy x-ray absorptiometry. Insulin resistance was measured by a hyperinsulinemic–euglycemic clamp.

Results. There was no significant difference between the IS and IR groups with respect to body mass index, body fat index, absolute amount of trunk fat, or percent body fat. However, the difference in LF/TF ratio between the IS (1.02 ± 0.05) and the IR groups (0.77 ± 0.05) was highly significantly different (p <.001). Insulin resistance had a stronger correlation to the LF/TF ratio (r = 0.61, p <.001) than to absolute trunk fat (r = –0.32, p =.051). Adiponectin levels had a strong association with the LF/TF ratio (r = 0.63, p <.001), but did not correlate to absolute trunk fat (r = –0.24, p =.18).

Conclusions. The distribution of body fat (LF/TF ratio) in elderly persons is a stronger determinant of insulin resistance and adiponectin levels than is trunk fat alone. The LF/TF ratio can be a useful tool to assess insulin sensitivity in the elderly population.

AGE is a strong risk factor for the development of insulin resistance. An understanding of the mechanisms leading to insulin resistance in the elderly population is critical to reducing the mortality, morbidity, and costs associated with clinical care associated with diabetes. In other conditions with a redistribution of adipose tissue from the peripheral to central sites such as congenital and acquired lipodystrophy syndromes, this distribution of body fat is strongly associated with insulin resistance (1,2). Lipodystrophy syndromes are characterized by loss of peripheral fat and accumulation of fat in the abdomen, liver, and skeletal muscle (1,3). Similarly, subcutaneous adipose tissue loss in the limbs is well known to accompany aging (4–8). It is intriguing that aging may represent another lipodystrophy syndrome with the loss of limb fat, increase in visceral fat, and insulin resistance. This study was undertaken to assess whether the relationship of limb fat to trunk fat (LF/TF ratio), as in lipodystrophy syndromes, is a stronger determinant of insulin resistance than limb fat or trunk fat independently (9). As in lipodystrophy syndromes, this identification of an inverse relationship between limb fat and trunk fat may focus areas of research on restoring peripheral fat loss in elderly persons as an indirect means of reducing visceral fat and insulin resistance (10).

Adipose tissue is known to secrete various hormones that may affect insulin sensitivity. Such hormones include adiponectin, which is thought to be important in signaling between adipose tissue, muscle, and liver tissue (11). Adiponectin levels are reduced in obese persons, type 2 diabetics, and in individuals with lipodystrophy syndromes, conditions which are all associated with insulin resistance (12). Therefore, the relationship of adiponectin levels to the LF/TF ratio was also investigated.


   MATERIALS AND METHODS
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Participants
Elderly individuals participated in this study, which was approved by the Committee on Research Involving Human Subjects at the Stony Brook University and Medical Center. All participants gave their written informed consent, and were 60 years old or older, healthy, and ambulatory. All participants had a body mass index (BMI) of 20–30 kg/m2. Exclusion criteria included an acute illness 3 months prior to enrollment, diagnosis of diabetes, or any medications that would affect insulin resistance. Such medications included glucocorticoids, nonsteroidal anti-inflammatory drugs, aspirin, and replacement hormones. A diagnosis of diabetes was based on self-report and fasting glucose levels.

Insulin Sensitivity
Insulin sensitivity was determined as the rate of infused glucose necessary to maintain euglycemia during infusion of insulin (assessed by a hyperinsulinemic–euglycemic clamp) (13). Participants were admitted to the General Clinical Research Center the night before the study, ate a uniform snack at 10 PM, and then fasted from midnight. At 7 AM after basal sampling, the participants were infused with 1.2 mU of insulin (Human; Eli Lilly, Indianapolis, IN)/(kg body weight/min) to elevate plasma insulin levels to ~ 40 µU/mL, sufficient to suppress hepatic glucose production in insulin-resistant states (14). Dextrose (10%) was administered intravenously at variable rates to maintain the plasma level at 90 mg/dL. Plasma glucose levels were assessed in arterialized blood samples obtained by the heated hand technique (15). Insulin sensitivity was determined between the second and third hour of insulin infusion. To normalize for differences in body composition, insulin sensitivity was expressed as milligrams of glucose per kilogram of lean body mass (LBM) as determined by dual-energy x-ray absorptiometry (DEXA).

Body Composition
Body composition, including LBM and total body fat, was determined by DEXA performed with a whole-body scanner (Hologic Inc., Bedford, MA). Limb fat was calculated as the total fat in arms and legs (g). In the upper arms, the limb fat included subcutaneous adipose tissue from the shoulder to the wrist. In the lower extremities, limb fat included subcutaneous adipose tissue from the hip to the ankle. Trunk fat was the amount of fat measured by the DEXA from below the neck to the pelvis. Trunk fat was measured as a surrogate for abdominal adiposity, as the DEXA scan software does not permit measurement of abdominal fat. More so, measuring the trunk fat from below the neck to the pelvis establishes easily identified and reproducible boundaries (neck and pelvis) between participants, whereas measuring abdominal fat may lead to variability between participants due to individual variation in abdominal shape. To correct for differences in height, body fat and trunk fat were also expressed as body fat mass and trunk fat mass indices, which were calculated by dividing these parameters by the square of the height (m2).

Biochemical Analysis
Plasma glucose levels were determined by the glucose oxidase method, using a Beckman Glucose Analyzer II (Brea, CA). High-density lipoprotein (HDL) was measured by enzymatic calorimetric assays performed on a Hitachi Modular Machine. Insulin levels were analyzed by radioimmunoassay (RIA; Diagnostic Products Corporation, Los Angeles, CA). C-reactive protein levels were analyzed by RIA. Serum adiponectin levels were measured by using a human enzyme-linked immunosorbent assay (ELISA) kit (LINCO Research, St. Charles, MO). Serum nonesterified fatty acids were measured by using the Wako NEFA C test kit (Wako Chemicals USA, Inc., Richmond, VA). The coefficient of variation within each assay was < 5%.

Statistical Analysis
All data are presented as means ± standard error of the mean. Differences between the insulin-sensitive and insulin-resistant participants were analyzed with the Student's t test. Pearson correlations were used to assess bivariate associations, and multiple regression modeling was used to assess multivariate associations and to identify factors predictive of insulin resistance. Differences were considered statistically significant if p <.05.


   RESULTS
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Whereas as a group elderly persons are known to be more insulin resistant than young persons, about half of elderly individuals may be insulin sensitive (16). Comparing participants with insulin resistance to those who were insulin sensitive controlled for any effects on metabolism that arose solely due to aging. Baseline characteristics of the participants are shown in Table 1. The participants were divided into an insulin sensitive (IS) and insulin resistant (IR) group. A glucose infusion rate of 8 mg glucose/kg LBM/min was used as a cutoff to divide the sample into an IS and IR group (8 mg glucose/kg LBM/min is the mean and median insulin sensitivity in the group). As illustrated in Table 1, the insulin sensitivity of the IS group is 10.40 ± 0.85, whereas that of the IR group is 5.68 ± 0.48 (p <.001). The IR group was ~5 years older (p =.01), confirming that even within a group of elderly persons, insulin resistance increases with increasing age. The body weight of the IR group was ~5 kg higher than that of the IS group; however, this difference did not reach statistical significance. As expected, fasting glucose levels were higher in the IR group; this finding indicates impaired fasting glucose by current American Diabetes Association criteria. Insulin levels were also higher in the IR group than in the IS group. In these participants, with a BMI of 20–30 kg/m2, the BMI was not significantly different between the IR and IS groups (Table 1). The amount of trunk fat was somewhat higher in the IR group than in the IS group (Table 1); however, when normalized for height (divided by height squared) there was no statistically significant difference. Body fat index (kg/height m2), trunk fat index (kg/m2), and percent body fat were not significantly different between the two groups (Table 1). Limb fat (arms and legs) was somewhat higher in the IS group compared to the IR group (Table 1), although this difference did not reach statistical significance. The distribution of fat, represented by LF/TF ratio, demonstrated the strongest difference between the IS (1.02 ± 0.05) and the IR group (0.77 ± 0.05, p <.001) (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1. Comparison of Body Fat Distribution in Insulin-Sensitive Versus Insulin-Resistant Elderly Persons.

 
Further analysis of the participants (IS and IR) as a group was performed with respect to insulin resistance. As previously described by others, the absolute amount of trunk fat was inversely correlated with insulin sensitivity (r = –0.32, p =.051) (17). However, the amount of limb fat was not significantly related to insulin sensitivity (r = 0.21, p =.20) (Figure 1A). In contrast, the LF/TF ratio was highly correlated with insulin sensitivity (r = 0.61, p <.001) (Figure 1B), to a larger extent than was absolute amount of trunk fat. To control for potential gender effects, a subgroup analysis of only elderly men also demonstrated the stronger correlation of insulin sensitivity with LF/TF ratio (r = 0.62, p =.001) than to trunk fat (r = –0.50, p =.01). Multivariate analysis revealed LF/TF ratio as the strongest determinant of insulin sensitivity in our participants (r = 0.56, p <.001), with age as the second strongest determinant (r = –0.34, p =.009).


Figure 01
View larger version (12K):
[in this window]
[in a new window]

  		Figure 1. Correlation of trunk fat (A) and limb fat/trunk fat ratio (B) with insulin sensitivity. Limb and trunk fat were quantified using dual-energy x-ray absorptiometry scan. Insulin sensitivity was assessed by the rate of glucose infusion required to maintain euglycemia during the period of 120–180 minutes of insulin infusion (1.2 mU/kg/min). LBM = lean body mass

 
The body fat distribution in all the participants (IS and IR combined) was also related to circulating levels of adiponectin. In the participants, the relationship of adiponectin levels to trunk fat (r = –0.24, p =.18) (Figure 2A) or to limb fat (r = 0.33, p =.07) were not statistically significant; however, adiponectin levels were highly correlated with the LF/TF ratio (r = 0.63, p <.001) (Figure 2B). The high correlation between adiponectin levels and LF/TF ratio was higher than the correlation between adiponectin levels and insulin sensitivity (r = 0.47, p =.003). Again, a subgroup analysis to control for gender indicated a strong correlation between adiponectin levels and the LF/TF ratio in elderly men (r = 0.77, p <.001), with no significant relationship of adiponectin levels to trunk fat in the elderly men (r = –0.31, p =.15). As dyslipidemia is associated with insulin resistance (18–20), the relationship between the LF/TF ratio and lipoprotein values was investigated. The LF/TF ratio was associated with HDL (r = 0.50, p <.01) and triglycerides (r = –0.30, p =.04), but not fatty acid levels (r = –0.03, p =.88).


Figure 02
View larger version (10K):
[in this window]
[in a new window]

  		Figure 2. Correlation of trunk fat (A) and limb fat/trunk fat ratio (B) with adiponectin levels. Limb and trunk fat were quantified using dual-energy x-ray absorptiometry scan

 

   DISCUSSION
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
In the present study, we show to our knowledge for the first time that the LF/TF ratio is a stronger determinant of insulin resistance than either limb fat or trunk fat independently. This information can be obtained by a simple calculation from a DEXA scan. We also show an association of LF/TF ratio to HDL levels. It is well known that women store fat more peripherally, and men store body fat more centrally, but the relationship of LF/TF to insulin sensitivity does not depend on gender. Men-only analysis still indicated the strong correlation between insulin sensitivity and LF/TF ratio.

Aging is well known to result in loss of subcutaneous fat (4–8,21). Prospective anthropometric measurements of elderly persons over a 6- to 10-year period have detected about a 17% decrease of skin-fold thickness in the extremities (4,7). In a comparison of healthy young and older men, computed tomography measurements revealed that elderly persons had a thigh subcutaneous fat area that was about half that of young men (156.3 ± 69.3 vs 82.4 ± 29.7 cm2) (5). Recent studies of elderly persons have also suggested a protective effect of subcutaneous adipose tissue of the limbs on the development of insulin resistance and cardiovascular diseases (18–20,22–24). In the Health, Aging and Body Composition study of elderly individuals, a larger amount of subcutaneous adipose tissue in the thigh (from computed tomography scan) was independently associated with lower glucose and lipid levels in men only (19). Similarly, in the Hoorn Study, a larger leg fat mass (from DEXA) in 275 men and 284 women was associated with lower glucose levels in women only (18). In contrast, multiple studies have demonstrated an independent effect of truncal fat and percent body fat on the development of insulin resistance in elderly persons (17,25–29). In a study of 166 healthy postmenopausal women, trunk fat was the strongest independent predictor of insulin resistance (20). In yet another study of older men and women, insulin resistance is more closely associated with abdominal adiposity than with age (17). Whereas previous studies have demonstrated independent effects of both trunk fat and limb fat on insulin sensitivity, this study demonstrates that combining peripheral (LF) and central (TF) substantially improves the ability of body composition to predict insulin sensitivity. The strong correlation of LF/TF ratio, rather than absolute limb fat or trunk fat, with insulin resistance suggests a new paradigm—that aging is a form of acquired lipodystrophy. This form of acquired lipodystrophy may be termed as ‘age-associated lipodystrophy.’

In addition to a relationship with insulin sensitivity, the LF/TF ratio is also strongly associated with the lipid-derived hormone, adiponectin. This hormone is inversely associated with insulin sensitivity, and is reduced in obese persons, type 2 diabetics, and individuals with congenital or acquired lipodystrophy syndrome (11,12,30,31). Similarly, in elderly persons, a decrease in adiponectin levels was associated with insulin resistance (11,28). Interestingly, the present study revealed that adiponectin levels did not correlate with trunk fat or limb fat independently, but were strongly correlated with the LF/TF ratio, suggesting a close link between peripheral and central fat distribution and adiponectin levels. The role of adiponectin and body fat distribution associated with aging needs to be further investigated.

Although the DEXA scan does not permit separate measurements of visceral and abdominal subcutaneous fat, or separate measurements of intermuscular, intramyocellular, and subcutaneous fat in the legs, it does provide potential clinically important information from a relatively simple procedure.

Conclusion
This work confirms by a more sensitive technique (hyperinsulinemic–euglycemic clamp) that, in elderly individuals, maintenance of body fat stores in the periphery is associated with maintaining insulin sensitivity. It also advances our understanding that the relationship of limb fat to trunk fat in the elderly population, expressed as LF/TF ratio, indicates a stronger relationship to both insulin resistance and to adiponectin levels, more so than does each adipose tissue depot independently. The correlation of the LF/TF ratio to insulin sensitivity persists even after adjusting for gender (men only). It provides a new paradigm that insulin resistance of aging is a form of an acquired lipodystrophy.


   Acknowledgments
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 
This work was supported by National Institutes of Health (NIH) Grant R01 AG17446-01A2 (to MAM), by NIH General Clinical Research Center Grant M01 RR10710-02 (Clinical Research Scholar Program, Center for Translational Research, School of Medicine, Health Sciences Center, State University of New York at Stony Brook; to SG), and by the Empire Clinical Research Investigators Program (to JJF and MMM).

We thank the participants in the study, Joyce Quick for her clinical assistance on the General Clinical Research Center, Jeanne Kidd for her help in coordinating this research, and the nursing and core laboratory staff of the General Clinical Research Center for their assistance.


   Footnotes
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 
Decision Editor: Luigi Ferrucci, MD, PhD

Received July 6, 2006

Accepted November 20, 2006


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. Garg A. Acquired and inherited lipodystrophies. N Engl J Med. 2004;350:1220-1234.[Free Full Text]
  2. Hadigan C, Meigs JB, Corcoran C, et al. Metabolic abnormalities and cardiovascular disease risk factors in adults with human immunodeficiency virus infection and lipodystrophy. Clin Infect Dis. 2001;32:130-139.[Medline]
  3. Grinspoon S. Insulin resistance in the HIV-lipodystrophy syndrome. Trends Endocrinol Metab. 2001;12:413-419.[Medline]
  4. Hughes VA, Roubenoff R, Wood M, Frontera WR, Evans WJ, Fiatarone Singh MA. Anthropometric assessment of 10-y changes in body composition in the elderly. Am J Clin Nutr. 2004;80:475-482.[Abstract/Free Full Text]
  5. Schwartz RS, Shuman WP, Bradbury VL, et al. Body fat distribution in healthy young and older men. J Gerontol. 1990;45:M181-M185.
  6. Shimokata H, Tobin JD, Muller DC, Elahi D, Coon PJ, Andres R. Studies in the distribution of body fat: I. Effects of age, sex, and obesity. J Gerontol. 1989;44:M66-M73.
  7. Chumlea WC, Rhyme RL, Garry PJ, Hunt WC. Changes in anthropometric indices of body composition with age in a healthy elderly population. Am J Hum Biol. 1989;1:457-462.
  8. Kirkland JL, Tchkonia T, Pirtskhalava T, Han J, Karagiannides I. Adipogenesis and aging: does aging make fat go MAD? Exp Gerontol. 2002;37:757-767.[Medline]
  9. Mynarcik DC, McNurlan MA, Steigbigel RT, Fuhrer J, Gelato MC. Association of severe insulin resistance with both loss of limb fat and elevated serum tumor necrosis factor receptor levels in HIV lipodystrophy. J Acquir Immune Defic Syndr. 2000;25:312-321.[Medline]
  10. Hadigan C, Yawetz S, Thomas A, Havers F, Sax PE, Grinspoon S. Metabolic effects of rosiglitazone in HIV lipodystrophy: a randomized, controlled trial. Ann Intern Med. 2004;140:786-794.[Abstract/Free Full Text]
  11. Ryan AS, Berman DM, Nicklas BJ, et al. Plasma adiponectin and leptin levels, body composition, and glucose utilization in adult women with wide ranges of age and obesity. Diabetes Care. 2003;26:2383-2388.[Abstract/Free Full Text]
  12. Gordon LB, Harten IA, Patti ME, Lichtenstein AH. Reduced adiponectin and HDL cholesterol without elevated C-reactive protein: clues to the biology of premature atherosclerosis in Hutchinson-Gilford Progeria Syndrome. J Pediatr. 2005;146:336-341.[Medline]
  13. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979;237:E214-E223.[Medline]
  14. Wise SD, Nielsen MF, Cryer PE, Rizza RA. Overnight normalization of glucose concentrations improves hepatic but not extrahepatic insulin action in subjects with type 2 diabetes mellitus. J Clin Endocrinol Metab. 1998;83:2461-2469.[Abstract/Free Full Text]
  15. Abumrad NN, Rabin D, Diamond MP, Lacy WW. Use of a heated superficial hand vein as an alternative site for the measurement of amino acid concentrations and for the study of glucose and alanine kinetics in man. Metabolism. 1981;30:936-940.[Medline]
  16. Cree MG, Newcomer BR, Katsanos CS, et al. Intramuscular and liver triglycerides are increased in the elderly. J Clin Endocrinol Metab. 2004;89:3864-3871.[Abstract/Free Full Text]
  17. Kohrt WM, Kirwan JP, Staten MA, Bourey RE, King DS, Holloszy JO. Insulin resistance in aging is related to abdominal obesity. Diabetes. 1993;42:273-281.[Abstract]
  18. Snijder MB, Dekker JM, Visser M, et al. Trunk fat and leg fat have independent and opposite associations with fasting and postload glucose levels: the Hoorn study. Diabetes Care. 2004;27:372-377.[Abstract/Free Full Text]
  19. Snijder MB, Visser M, Dekker JM, et al. Low subcutaneous thigh fat is a risk factor for unfavourable glucose and lipid levels, independently of high abdominal fat. The Health ABC Study. Diabetologia. 2005;48:301-308.[Medline]
  20. Van Pelt RE, Evans EM, Schechtman KB, Ehsani AA, Kohrt WM. Contributions of total and regional fat mass to risk for cardiovascular disease in older women. Am J Physiol Endocrinol Metab. 2002;282:E1023-E1028.[Abstract/Free Full Text]
  21. Visser M, Pahor M, Tylavsky F, et al. One- and two-year change in body composition as measured by DXA in a population-based cohort of older men and women. J Appl Physiol. 2003;94:2368-2374.[Abstract/Free Full Text]
  22. Tanne D, Medalie JH, Goldbourt U. Body fat distribution and long-term risk of stroke mortality. Stroke. 2005;36:1021-1025.[Abstract/Free Full Text]
  23. Hara M, Saikawa T, Kurokawa M, Sakata T, Yoshimatsu H. Leg fat percentage correlates negatively with coronary atherosclerosis. Circ J. 2004;68:1173-1178.[Medline]
  24. Permana PA, Nair S, Lee YH, Luczy-Bachman G, Vozarova De Courten B, Tataranni PA. Subcutaneous abdominal preadipocyte differentiation in vitro inversely correlates with central obesity. Am J Physiol Endocrinol Metab. 2004;286:E958-E962.[Abstract/Free Full Text]
  25. Ozbey N, Sencer E, Molvalilar S, Orhan Y. Body fat distribution and cardiovascular disease risk factors in pre- and postmenopausal obese women with similar BMI. Endocr J. 2002;49:503-509.[Medline]
  26. Barbieri M, Rizzo MR, Manzella D, Paolisso G. Age-related insulin resistance: is it an obligatory finding? The lesson from healthy centenarians. Diabetes Metab Res Rev. 2001;17:19-26.[Medline]
  27. Basu R, Breda E, Oberg AL, et al. Mechanisms of the age-associated deterioration in glucose tolerance: contribution of alterations in insulin secretion, action, and clearance. Diabetes. 2003;52:1738-1748.[Abstract/Free Full Text]
  28. Cnop M, Havel PJ, Utzschneider KM, et al. Relationship of adiponectin to body fat distribution, insulin sensitivity and plasma lipoproteins: evidence for independent roles of age and sex. Diabetologia. 2003;46:459-469.[Medline]
  29. Coon PJ, Rogus EM, Drinkwater D, Muller DC, Goldberg AP. Role of body fat distribution in the decline in insulin sensitivity and glucose tolerance with age. J Clin Endocrinol Metab. 1992;75:1125-1132.[Abstract]
  30. Haque WA, Shimomura I, Matsuzawa Y, Garg A. Serum adiponectin and leptin levels in patients with lipodystrophies. J Clin Endocrinol Metab. 2002;87:2395-2398.[Abstract/Free Full Text]
  31. Mynarcik DC, Combs T, McNurlan MA, Scherer PE, Komaroff E, Gelato MC. Adiponectin and leptin levels in HIV-infected subjects with insulin resistance and body fat redistribution. J Acquir Immune Defic Syndr. 2002;31:514-520.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
PubMed
Right arrow PubMed Citation

HOME ARCHIVE SEARCH TABLE OF CONTENTS